How to find the coordinates from an convolution with an mask

Hi,

I created an rectagular mask or elliptical mask that i rotate and convolute with an gray-scale image with some intensity content. I like to plot the position of the mask (mask that have the highest intensity value), but i am having a little trouble to figure out how to find the coordiantes of the mask that gave the highest intensity so i can then mark out the location of the worst-case rectangular or elliptical mask with an line contour. Any ideas?

Thanks!

Answer 1

[Dhruv]

To find the coordinates of the mask that gives the highest intensity, you can use the following steps:

• Convolve the image with the mask. This will create a new image, called the convolved image, where each pixel value is the sum of the product of the mask values and the image values under the mask.
• Find the maximum intensity value in the convolved image using the “max” function (Refer to the documentation: https://www.mathworks.com/help/releases/R2021b/matlab/ref/max.html)
• Use the “find” function to find the coordinates of all the pixels in the convolved image that have the maximum intensity value. (Refer to the documentation: https://www.mathworks.com/help/releases/R2021b/matlab/ref/find.html)

Here is an example code:

function find_max_intensity_coordinates(image, mask)
  % convolve the image with the mask
  convolved_image = conv2(image, mask)

  % find the maximum intensity value in the convolved image
  max_intensity = max(convolved_image)

  % find the coordinates of the maximum intensity value
  max_intensity_coordinates = find(convolved_image == max_intensity)

  return max_intensity_coordinates

The function returns the coordinates of the maximum intensity pixels. Further, you may mark it with a line contour or any other desired visualization.

I hope this helps!